The overriding problem of understanding and controlling blood-surface reactions is inability to relate precise but limited information obtained from in vitro studies to the process that takes place when a cardiovascular prosthesis or artificial organ is employed. Work proposed here would link to clinical conditions well-established descriptions of how blood proteins in unmoving, highly dilute plasma solutions adsorb, react and desorb at biomaterial surfaces. The proposed work involves four capabilities: (1) Establishing "separated flows" in a convenient laboratory environment. (Within such flows, at the start of clinical procedures, whole blood is likely to be diluted before it contacts biomaterial surfaces, and such flows are suspected of favoring the initiation of thrombosis.) (2) Imaging patterns of protein deposition on surfaces which bound separated flows, utilizing metal oxide particles that have been rendered specific to designated proteins. (3) Using biological tests, in situ, for determining the reactivity of the adsorbed proteins. (4) Analyzing the geometric aspects of these patterns using a scanning device and computer workstation for image analysis. Conclusions drawn from this work would be confirmed by contacting surfaces with whole blood in the same separated-flow apparatus. Surfaces to be used include cleaned glass as well as the NHLBI reference biomaterials, polydimethylsiloxane and low-density polyethylene.